We have an ongoing project to characterize the antigens of mouse, rat, and human Pneumocystis. We have previously purified the major surface glycoprotein (MSG) of both rat and human pneumocystis using HPLC. It is necessary to use Pneumocystis from both sources because the organisms are different species. Subsequently, we identified a number of clones from a cDNA library of rat Pneumocystis that contain genes encoding for the MSG. These clones were clearly related but not identical, demonstrating that multiple genes encode the MSG. We have continued studies to characterize potential antigens of Pneumocystis. We have cloned a number of human Pneumocystis MSG genes and have expressed a full-length MSG in two fragments. We developed an ELISA to examine antibody responses to these antigens, and have utilized it to examine sera from patients with or without HIV infection, and with or without a history of PCP, as well as sera from a variety of healthy controls. In about 15% of healthy patients followed serially we have been able to document changes in antibody titers, suggesting that these individuals have developed reinfection or reactivation of Pneumocystis infection. We will continue these studies to better understand the epidemiology of Pneumocystis infection in humans. We have also identified the unique expression site of MSG in human Pneumocystis, and can now identify the MSG variants that are expressed in a patient with PCP. Within this expression site we have identified a region of tandem repeats that varies among different Pneumocystis isolates, and thus provides a method for typing human Pneumocystis. We have sequenced MSG variants from a single human pneumocystis isolate and have shown that the isolates can be divided, into 2 subgroups that are genetically distinct. We have cloned and sequenced MSG variants from additional patients from around the world to see if this will provide information about the distribution of different Pneumocystis strains. To date we have found substantial variability in MSG variants from different parts of the world, but no clustering by geographic location. To better examine the MSG repertoire and MSG variability, we have utilized next generation sequencing. By 454 sequencing, we were unable to reconstruct the original sequences using a mixture of plasmids, in large part because of the conserved regions in MSG variants. PacBio sequencing, which allows potentially longer reads, appears to be able to overcome the assembly problem. We are continuing to evaluate results generated by PacBio sequencing using mouse and human isolates of Pneumocystis. We are also examining MSG variants from Pneumocystis infecting other host species, to develop a better picture of the evolution of MSG variants. We have also expressed a p57 protein of P. murina that is predicted to be a surface protein based on a predicted GPI-type anchor. Based on western blot studies this is an antigen recognized by mouse serum from exposed or infected animals. By immunohistochemical labeling, we have been able to demonstrate that p57 appears to be a stage-specific antigen that is expressed by intracystic bodies as well as small trophic forms. Further, organisms typically express either p57 or MSG, but not both together. This suggests that p57 replaces MSG during intracystic development, where the antigenic variation conferred by MSG is not necessary. We have sequenced the genomes of 3 Pneumocystis species, and are using the data from the genomes to identify potential surface proteins that may be antigens, based on the predicted amino acid sequence characteristics. The goal of this study is to better understand the pathogenesis of Pneumocystis pneumonia with the hope that we can use this information to control or prevent this disease.